KR102424755B1 - High-band signal modeling - Google Patents

Abstract

The method includes, at a speech encoder, filtering an audio signal into a first group of sub-bands within a first frequency range and a second group of sub-bands within a second frequency range. The method also includes generating a harmonically extended signal based on the first group of sub-bands. The method further includes generating a third group of sub-bands based at least in part on the harmonically extended signal. The third group of sub-bands corresponds to the second group of sub-bands. The method also includes determining a first adjustment parameter for a first sub-band in the third group of sub-bands or a second adjustment parameter for a second sub-band in the third group of sub-bands do. The first adjustment parameter is based on a metric of the first sub-band in the second group of sub-bands, and the second adjustment parameter is based on a metric of the second sub-band in the second group of sub-bands.

Description

High-Band Signal Modeling {HIGH-BAND SIGNAL MODELING}

claim of priority

This application claims priority to U.S. Patent Application No. 14/568,359, filed December 12, 2014, and U.S. Provisional Application No. 61/916,697, filed December 16, 2013, both of which are "HIGH-BAND SIGNAL MODELING", the contents of which are incorporated by reference in their entirety.

technical field

BACKGROUND This disclosure relates generally to signal processing.

Advances in technology have resulted in smaller and more powerful computing devices. For example, a variety of portable personal computing devices are currently available, including wireless computing devices such as portable wireless telephones, personal digital assistants (PDAs), and paging devices that are small, lightweight, and easy for users to carry. exist. More specifically, portable wireless telephones, such as cellular telephones and Internet protocol (IP) telephones, are capable of communicating voice and data packets over wireless networks. Also, many of these wireless telephones include other types of devices included therein. For example, wireless telephones also include digital still cameras, digital video cameras, digital recorders, and audio file players.

In conventional telephone systems (eg, public switched telephone networks (PSTNs)), the signal bandwidth is limited to a frequency range of 300 Hertz (Hz) to 3.4 kiloHertz (kHz). In wideband (WB) applications, such as cellular telephone and voice over internet protocol (VoIP), the signal bandwidth may span a frequency range from 50 Hz to 7 kHz. Super wideband (SWB) coding schemes support a bandwidth that extends up to about 16 kHz. Extending the signal bandwidth from narrowband telephony at 3.4 kHz to SWB telephony at 16 kHz may improve the quality of signal recovery, intelligibility, and naturalness.

SWB coding techniques typically involve encoding and transmitting a low frequency portion of a signal (eg, 50 Hz to 7 kHz, also referred to as “low-band”). For example, the low-band may be represented using filter parameters and/or a low-band excitation signal. However, to improve coding efficiency, the high frequency portion of the signal (eg, 7 kHz to 16 kHz, also referred to as “high-band”) may not be fully encoded and transmitted. Instead, the receiver may utilize signal modeling to predict the high-band. In some implementations, data associated with the high-band may be provided to a receiver to aid in prediction. Such data may be referred to as “side information” and may include gain information, line spectral frequencies (line spectral frequencies (LSFs), also referred to as line spectral pairs (LSPs)), and the like. Properties of the low-band signal may be used to generate side information; Energy gaps between the low-band and the high-band may result in additional information that incorrectly characterizes the high-band.

Systems and methods for performing highband signal modeling are disclosed. A first filter (eg, a quadrature mirror filter (QMF) bank or pseudo-QMF bank) converts the audio signal to a first group of sub-bands corresponding to a low-band portion of the audio signal and a high-band portion of the audio signal. may filter into a second group of corresponding sub-bands corresponding to . The group of sub-bands corresponding to the low band portion of the audio signal and the group of sub-bands corresponding to the high band portion of the audio signal may or may not have common sub-bands. The synthesis filter bank may combine the first group of sub-bands to generate a low-band signal (eg, a low-band residual signal), which may be provided to the low-band coder. The low-band coder may quantize the low-band signal using a Linear Prediction Coder (LP Coder), which may generate a low-band excitation signal. The non-linear transformation process may generate a harmonically extended signal based on the low-band excitation signal. The bandwidth of the non-linear excitation signal may be larger than the low band portion of the audio signal or may be on the same order as the low band portion of the entire audio signal. For example, the non-linear transform generator may up-sample the low-band excitation signal, and process the up-sampled signal through a non-linear function to generate harmonics having a bandwidth greater than the bandwidth of the low-band excitation signal. It is also possible to generate an extended signal.

In a particular embodiment, the second filter may split the harmonically extended signal into a plurality of sub-bands. In this embodiment, modulated noise is added to each sub-band of the plurality of sub-bands of the harmonically extended signal to form sub-bands (eg, a sub-band corresponding to the high-band of the harmonically extended signal). a third group of sub-bands corresponding to the second group of bands). In another particular embodiment the modulated noise may be mixed with the harmonically extended signal to produce a high-band excitation signal that is provided to a second filter. In such an embodiment, the second filter may split the high-band excitation signal into a third group of sub-bands.

The first parameter estimator may determine a first adjustment parameter for the first sub-band in the third group of sub-bands based on a metric of the corresponding sub-band in the second group of sub-bands. For example, the first parameter estimator may determine a spectral relationship and/or a temporal envelope relationship between the first sub-band in the third group of sub-bands and a corresponding high-band portion of the audio signal. In a similar manner, the second parameter estimator determines a second adjustment parameter for the second sub-band in the third group of sub-bands based on the metric of the corresponding sub-band in the second group of sub-bands. may be The adjustment parameters may be quantized and transmitted along with other side information to the decoder to assist the decoder in reconstructing the high-band portion of the audio signal.

In a particular aspect, a method includes, at a speech encoder, filtering an audio signal into a first group of sub-bands within a first frequency range and a second group of sub-bands within a second frequency range. The method also includes generating a harmonically extended signal based on the first group of sub-bands. The method further includes generating a third group of sub-bands based at least in part on the harmonically extended signal. The third group of sub-bands corresponds to the second group of sub-bands. The method also includes determining a first adjustment parameter for a first sub-band in the third group of sub-bands or a second adjustment parameter for a second sub-band in the third group of sub-bands do. The first adjustment parameter is based on a metric of the first sub-band in the second group of sub-bands, and the second adjustment parameter is based on a metric of the second sub-band in the second group of sub-bands.

In another particular aspect, an apparatus includes a first filter configured to filter an audio signal into a first group of sub-bands within a first frequency range and a second group of sub-bands within a second frequency range. The apparatus also includes a non-linear transform generator configured to generate a harmonically extended signal based on the first group of sub-bands. The apparatus further includes a second filter configured to generate a third group of sub-bands based at least in part on the harmonically extended signal. The third group of sub-bands corresponds to the second group of sub-bands. The apparatus is further configured to determine a parameter estimator configured to determine a first adjustment parameter for a first sub-band in the third group of sub-bands or a second adjustment parameter for a second sub-band in the third group of sub-bands. include those The first adjustment parameter is based on a metric of the first sub-band in the second group of sub-bands, and the second adjustment parameter is based on a metric of the second sub-band in the second group of sub-bands.

In another particular aspect, a non-transitory computer-readable medium, when executed by a processor in a speech encoder, causes the processor to: convert an audio signal to a first group of sub-bands within a first frequency range and a sub-band within a second frequency range. - contains instructions to filter into the second group of bands. The instructions are also executable to cause the processor to generate a harmonically extended signal based on the first group of sub-bands. The instructions are further executable to cause the processor to generate a third group of sub-bands based at least in part on the harmonically extended signal. The third group of sub-bands corresponds to the second group of sub-bands. The instructions also cause the processor to: configure a first adjustment parameter for a first sub-band in the third group of sub-bands or a second adjustment parameter for a second sub-band in the third group of sub-bands. actionable to make a decision. The first adjustment parameter is based on a metric of the first sub-band in the second group of sub-bands, and the second adjustment parameter is based on a metric of the second sub-band in the second group of sub-bands.

In another particular aspect, an apparatus includes means for filtering an audio signal into a first group of sub-bands within a first frequency range and a second group of sub-bands within a second frequency range. The apparatus also includes means for generating a harmonically extended signal based on the first group of sub-bands. The apparatus further includes means for generating a third group of sub-bands based at least in part on the harmonically extended signal. The third group of sub-bands corresponds to the second group of sub-bands. The apparatus also includes means for determining a first adjustment parameter for a first sub-band in the third group of sub-bands or a second adjustment parameter for a second sub-band in the third group of sub-bands do. The first adjustment parameter is based on a metric of the first sub-band in the second group of sub-bands, and the second adjustment parameter is based on a metric of the second sub-band in the second group of sub-bands.

In another particular aspect, a method includes generating, at a speech decoder, a harmonically extended signal based on a low-band excitation signal generated by a linear prediction based decoder based on parameters received from the speech encoder. The method further includes generating the group of high-band excitation sub-bands based at least in part on the harmonically extended signal. The method also includes adjusting the group of high-band excitation sub-bands based on the adjustment parameters received from the speech encoder.

In other particular aspects, an apparatus includes a non-linear transform generator configured to generate a harmonically extended signal based on a low-band excitation signal generated by a linear prediction based decoder based on parameters received from a speech encoder. . The apparatus further includes a second filter configured to generate the group of high-band excitation sub-bands based at least in part on the harmonically extended signal. The apparatus also includes adjusters configured to adjust the group of high-band excitation sub-bands based on the adjustment parameters received from the speech encoder.

In another particular aspect, an apparatus includes means for generating a harmonically extended signal based on a low-band excitation signal generated by a linear prediction based decoder based on parameters received from a speech encoder. The apparatus further includes means for generating the group of high-band excitation sub-bands based at least in part on the harmonically extended signal. The apparatus also includes means for adjusting the group of high-band excitation sub-bands based on the adjustment parameters received from the speech encoder.

In another particular aspect, a non-transitory computer-readable medium, when executed by a processor at a speech decoder, causes the processor to: low-band excitation generated by a linear prediction based decoder based on parameters received from the speech encoder. and instructions for generating a harmonically extended signal based on the signal. The instructions are further executable to cause the processor to generate the group of high-band excitation sub-bands based at least in part on the harmonically extended signal. The instructions are also executable to cause the processor to adjust the group of high-band excitation sub-bands based on the adjustment parameters received from the speech encoder.

Certain advantages provided by at least one of the disclosed embodiments include improved resolution modeling of a high-band portion of an audio signal. Other aspects, advantages, and features of the present disclosure will become apparent after review of the entire application, including the following sections: Brief Description of the Drawings, Detailed Description of the Invention, and the Claims.

1 is a diagram illustrating a particular embodiment of a system operable to perform high-band signal modeling;
2 is a diagram illustrating another specific embodiment of a system operable to perform high-band signal modeling;
3 is a diagram illustrating another specific embodiment of a system operable to perform high-band signal modeling;
4 is a diagram of a particular embodiment of a system operable to restore an audio signal using adjustment parameters;
5 is a flowchart of a particular embodiment of a method for performing high-band signal modeling;
6 is a flowchart of a particular embodiment of a method of reconstructing an audio signal using adjustment parameters; and
7 is a block diagram of a wireless device operable to perform signal processing operations in accordance with the systems and methods of FIGS. 1-6 ;

1 , a particular embodiment of a system operable to perform high-band signal modeling is shown and generally designated 100 . In a particular embodiment, system 100 may be integrated into an encoding system or apparatus (eg, into a wireless telephone or coder/decoder (codec)). In other embodiments, system 100 may be integrated into a set top box, music player, video player, entertainment unit, navigation device, communication device, PDA, fixed location data unit, or computer.

It should be noted that, in the following description, various functions performed by the system 100 of FIG. 1 are described as being performed by certain components or modules. However, this division of components and modules is for illustrative purposes only. In an alternative embodiment, the functionality performed by a particular component or module may instead be partitioned among multiple components or modules. Also, in an alternative embodiment, two or more components or modules of FIG. 1 may be integrated into a single component or module. Each component or module shown in FIG. 1 includes hardware (eg, a field-programmable gate array (FPGA) device, an application-specific integrated circuit (ASIC), a digital signal processor ( digital signal processor (DSP) controller, etc.), software (eg, instructions executable by a processor), or any combination thereof.

System 100 includes a first analysis filter bank 110 (eg, a QMF bank or pseudo-QMF bank) configured to receive an input audio signal 102 . For example, the input audio signal 102 may be provided by a microphone or other input device. In a particular embodiment, the input audio signal 102 may include speech. The input audio signal 102 may be a SWB signal that includes data in a frequency range from about 50 Hz to about 16 kHz. The first analysis filter bank 110 may filter the input audio signal 102 into multiple portions based on frequency. For example, first analysis filter bank 110 may generate a first group 122 of sub-bands within a first frequency range and a second group 124 of sub-bands within a second frequency range. The first group of sub-bands 122 may include M sub-bands, where M is an integer greater than zero. The second group of sub-bands 124 may include N sub-bands, where N is an integer greater than one. Accordingly, the first group of sub-bands 122 may include at least one sub-band, and the second group of sub-bands 124 includes two or more sub-bands. In certain embodiments, M and N may be similar values. In another particular embodiment, M and N may be different values. The first group of sub-bands 122 and the second group of sub-bands 124 may have the same or unequal bandwidth, and may or may not overlap. In an alternative embodiment, first analysis filter bank 110 may generate more than two groups of sub-bands.

The first frequency range may be lower than the second frequency range. In the example of FIG. 1 , the first group of sub-bands 122 and the second group of sub-bands 124 occupy non-overlapping frequency bands. For example, the first group of sub-bands 122 and the second group of sub-bands 124 may occupy non-overlapping frequency bands of 50 Hz - 7 kHz and 7 kHz - 16 kHz, respectively. In an alternative embodiment, the first group of sub-bands 122 and the second group of sub-bands 124 may occupy non-overlapping frequency bands of 50 Hz - 8 kHz and 8 kHz - 16 kHz, respectively. have. In another alternative embodiment, the first group of sub-bands 122 and the second group of sub-bands 124 overlap (eg, 50 Hz - 8 kHz and 7 kHz - 16 kHz, respectively), This may enable the low-pass filter and the high-pass filter of the first analysis filter bank 110 to have smooth rolloff, which simplifies the design and reduces the efficiency of the low-pass filter and the high-pass filter of the first analysis filter bank 110 . It can also reduce costs. The overlap of the first group of sub-bands 122 and the second group of sub-bands 124 may also enable smooth blending of the low-band signal and the high-band signal at the receiver, which in less It may result in audible artifacts.

It should be noted that although the example of FIG. 1 shows the processing of a SWB signal, this is for illustrative purposes only. In an alternative embodiment, the input audio signal 102 may be a WB signal having a frequency range of about 50 Hz to about 8 kHz. In such an embodiment, the first group of sub-bands 122 may correspond to a frequency range of about 50 Hz to about 6.4 kHz, and the second group of sub-bands 124 may correspond to a frequency range of about 6.4 kHz to about 8 kHz. It may correspond to a frequency range.

The system 100 may include a low-band analysis module 130 configured to receive the first group 122 of sub-bands. In a particular embodiment, the low-band analysis module 130 may represent an embodiment of a code excited linear prediction (CELP) encoder. The low-band analysis module 130 may include a linear prediction (LP) analysis and coding module 132 , a linear prediction coefficient (LPC) to LPS transform module 134 , and a quantizer 136 . LSPs may also be referred to as LSFs, and the two terms (LSP and LSF) may be used interchangeably herein. The LP analysis and coding module 132 may encode the spectral envelope of the first group of sub-bands 122 as a set of LPCs. The LPCs are for each frame of audio (eg, 20 milliseconds (ms) of audio corresponding to 320 samples at a sampling rate of 16 kHz), each sub-frame of audio (eg, 5 of audio ms), or any combination thereof. The number of LPCs generated for each frame or sub-frame may be determined by the “order” of the LP analysis performed. In a particular embodiment, the LP analysis and coding module 132 may generate a set of 11 LPCs corresponding to the tenth-order LP analysis.

The LPC-to-LSP transform module 134 may transform the set of LPCs generated by the LP analysis and coding module 132 (eg, using a one-to-one transform) into a corresponding set of LSPs. . Alternatively, the set of LPCs corresponds to parcor coefficients, log-area-ratio values, immittance spectral pairs (ISP), or immittance spectral frequencies (ISF). A set may be converted one-to-one. The transformation between the set of LPCs and the set of LSPs may be reversible without error.

The quantizer 136 may quantize the set of LSPs generated by the LPC-to-LSP transform module 134 . For example, the quantizer 136 may include or be coupled to multiple codebooks that include multiple entries (eg, vectors). To quantize the set of LSPs, the quantizer 136 may identify entries in codebooks that are “closest” to the set of LSPs (eg, based on a distortion measure such as least squares method or mean square error). . Quantizer 136 may output an index value or series of index values corresponding to the location of the identified entries in the codebook. The output of quantizer 136 thus represents the low-band filter parameters included in low-band bit stream 142 .

The low-band analysis module 130 may also generate a low-band excitation signal 144 . For example, low-band excitation signal 144 may be an encoded signal generated by coding an LP residual signal generated during an LP process performed by low-band analysis module 130 .

The system 100 is configured to receive a second group of sub-bands 124 from a first analysis filter bank 110 and a low-band excitation signal 144 from a low-band analysis module 130 . It may further include a module 150 . The high-band analysis module 150 may generate the high-band side information 172 based on the second group of sub-bands 124 and the low-band excitation signal 144 . For example, high-band side information 172 may include high-band LPCs and/or gain information (eg, adjustment parameters).

The high-band analysis module 150 may include a non-linear transform generator 190 . The non-linear transform generator 190 may be configured to generate a harmonically extended signal based on the low-band excitation signal 144 . For example, the non-linear transform generator 190 may up-sample the low-band excitation signal 144 , and process the up-sampled signal with a non-linear function to process the low-band excitation signal 144 . ) may generate a harmonically extended signal having a bandwidth greater than the bandwidth of .

The high-band analysis module 150 may also include a second analysis filter bank 192 . In a particular embodiment, the second analysis filter bank 192 may split the harmonically extended signal into a plurality of sub-bands. In this embodiment, modulated noise is added to each sub-band of the plurality of sub-bands of the harmonically extended signal to add a third group of sub-bands 126 corresponding to the second group of sub-bands 124 . ) (eg, high-band excitation signals). As non-limiting examples, the first sub-band H1 of the second group of sub-bands 124 may have a bandwidth ranging from 7 kHz to 8 kHz, and the second group of sub-bands 124 . The second sub-band H2 of ) may have a bandwidth ranging from 8 kHz to 9 kHz. Similarly, the first sub-band (not shown) of the third group of sub-bands 126 (corresponding to the first sub-band H1 ) may have a bandwidth ranging from 7 kHz to 8 kHz. and a second sub-band (not shown) of the third group of sub-bands 126 (corresponding to the second sub-band H2) may have a bandwidth ranging from 8 kHz to 9 kHz . In another particular embodiment, the modulated noise may be mixed with the harmonically extended signal to produce a high-band excitation signal provided to the second analysis filter bank 192 . In this embodiment, the second analysis filter bank 192 may split the high-band excitation signal into the third group of sub-bands 126 .

The parameter estimators 194 in the high-band analysis module 150 are configured to calculate the metric in the third group of sub-bands 126 based on the metric of the corresponding sub-band in the second group 124 of sub-bands. A first adjustment parameter (eg, an LPC adjustment parameter and/or a gain adjustment parameter) for the first sub-band may be determined. For example, the particular parameter estimator may include a first sub-band in the third group of sub-bands 126 and a corresponding high-band portion of the input audio signal 102 (eg, a second of sub-bands). The spectral relationship and/or envelope relationship between corresponding sub-bands in group 124 may be determined. In a similar manner, another parameter estimator is a second parameter estimator for the second sub-band in the third group of sub-bands 126 based on the metric of the corresponding sub-band in the second group of sub-bands 124 . 2 Adjustment parameters may be determined. As used herein, a “metric” of a sub-band may correspond to any value that characterizes the sub-band. As non-limiting examples, the metric of the sub-band may correspond to the signal energy of the sub-band, the residual energy of the sub-band, LP coefficients of the sub-band, and the like.

In a particular embodiment, the parameter estimators 194 include sub-bands (eg, components of the high-band portion of the input audio signal 102 ) of the second group of sub-bands 124 and a sub-band at least two gain factors (eg, adjustment parameters) according to a relationship between the corresponding sub-bands (eg, components of the high-band excitation signal) of the third group of bands 126 . can also be calculated. The gain factors may correspond to a difference (or ratio) between the energies of the corresponding sub-bands over a frame or some portion of a frame. For example, parameter estimators 194 may calculate the energy as a sum of squares of samples of each sub-frame for each sub-band, and the gain factor for each sub-frame is a ratio of those energies. may be the square root of In another particular embodiment the parameter estimators 194 calculate a time-varying relationship between the sub-bands of the second group of sub-bands 124 and the corresponding sub-bands of the third group of sub-bands 126 . The gain envelope may be calculated accordingly. However, the temporal envelope of the high-band portion of the input audio signal 102 (eg, the high-band signal) and the temporal envelope of the high-band excitation signal are likely to be similar.

In another particular embodiment, the parameter estimators 194 may include an LP analysis and coding module 152 and an LPC to LSP transform module 154 . Each of the LP analysis and coding module 152 and the LPC to LSP transform module 154 refers to corresponding components of the low-band analysis module 130 , but (eg, more for each coefficient, LSP, etc.) It may function as described above with a relatively reduced resolution (using fewer bits). LP analysis and coding module 152 may generate a set of LPCs that are transformed into LSPs by transform module 154 and quantized by quantizer 156 based on codebook 163 . For example, the LP analysis and coding module 152 , the LPC to LSP transform module 154 , and the quantizer 156 use the second group of sub-bands 124 to use the high-band filter information (e.g., For example, high-band LSPs or adjustment parameters) and/or high-band gain information included in high-band side information 172 may be determined.

The quantizer 156 may be configured to quantize the adjustment parameters from the parameter estimators 194 as high-band side information 172 . The quantizer may also be configured to quantize a set of spectral frequency values, such as LSPs, provided by transform module 154 . In other embodiments, quantizer 156 may receive and quantize sets of one or more other types of spectral frequency values in addition to or instead of LSFs or LSPs. For example, the quantizer 156 may receive and quantize the set of LPCs generated by the LP analysis and coding module 152 . Other examples include sets of Parco coefficients, log-area-ratio values, and ISFs that may be received and quantized at quantizer 156 . Quantizer 156 may include a vector quantizer that encodes an input vector (eg, a set of spectral frequency values in a vector format) into an index to a corresponding entry in a codebook or table, such as codebook 163 . . As another example, the quantizer 156 may be configured to determine one or more parameters from which the input vector may be dynamically generated at the decoder, such as in a sparse codebook embodiment, rather than retrieved from storage. To illustrate, the sparsity codebook examples may be applied to coding techniques such as CELP and codecs that conform to industry standards such as Third Generation Partnership 2 (3GPP2) Enhanced Variable Rate Codec (EVRC). In another embodiment, high-band analysis module 150 may include a quantizer 156 that generates synthesized signals (eg, according to a set of filter parameters) using multiple codebook vectors and , may be configured to select one of the codebook vectors associated with the synthesized signal that best matches the second group of sub-bands 124 , such as in the perceptually weighted domain.

In a particular embodiment, high-band side information 172 may include high-band LSPs as well as high-band gain parameters. For example, high-band side information 172 may include adjustment parameters generated by parameter estimators 194 .

The low-band bit stream 142 and the high-band side information 172 may be multiplexed by a multiplexer (MUX) 170 to produce an output bit stream 199 . The output bit stream 199 may represent an encoded audio signal corresponding to the input audio signal 102 . For example, the multiplexer 170 inserts the adjustment parameters included in the high-band side information 172 into the encoded version of the input audio signal 102 to adjust the gain (e.g., For example, it may be configured to enable envelope-based adjustment) and/or linearity adjustment (eg, spectrum-based adjustment). The output bit stream 199 may be transmitted (eg, via a wired, wireless, or optical channel) and/or stored by the transmitter 198 . At the receiver, inverse operations are performed by a demultiplexer (DEMUX), a low-band decoder, a high-band decoder, and a filter bank to convert an audio signal (eg, an input audio signal 102 provided to a speaker or other output device). restored version) can also be created. The number of bits used to represent the low-band bit stream 142 may be substantially greater than the number of bits used to represent the high-band side information 172 . Accordingly, most of the bits in output bit stream 199 may represent low-band data. The high-band side information 172 may be used at the receiver to regenerate a high-band excitation signal from the low-band data according to a signal model. For example, the signal model may be between low-band data (eg, a first group of sub-bands 122 ) and high-band data (eg, a second group of sub-bands 124 ). It may indicate an expected set of relationships or correlations. Thus, different signal models may be used for different kinds of audio data (eg, speech, music, etc.), and the particular signal model in use may be negotiated by the transmitter and receiver prior to communication of the encoded audio data. (or it may be defined by industry standards). Using the signal model, it may be possible for the high-band analysis module 150 at the transmitter to generate the high-band side information 172 so that the corresponding high-band analysis module at the receiver uses the signal model to output the bit stream. A second group of sub-bands 124 can be recovered from (199).

The system 100 of FIG. 1 includes synthesized high-band signal components (eg, a third group of sub-bands 126 ) and original high-band signal components (eg, of sub-bands). The degree of correlation between the second group 124 may be improved. For example, the spectral and envelope approximation between the synthesized high-band signal components and the original high-band signal components is a sub-band by sub-band basis for the metrics of the second group of sub-bands 124 of the sub-bands. It may be performed at a “finer” level by comparing to the third group of metrics 126 . The third group of sub-bands 126 may be adjusted based on adjustment parameters resulting from the comparison, which may be transmitted to a decoder to reduce audible artifacts during high-band reconstruction of the input audio signal 102 . have.

Referring to FIG. 2 , a particular embodiment of a system 200 operable to perform high-band signal modeling is shown. The system 200 includes a first analysis filter bank 110 , an analysis filter bank 202 , a low-band coder 204 , a non-linear transform generator 190 , a noise combiner 206 , a second analysis filter bank ( 192), and N parameter estimators 294a-294c.

The first analysis filter bank 110 may receive the input audio signal 102 and may be configured to filter the input audio signal 102 into multiple portions based on frequency. For example, first analysis filter bank 110 may generate a first group 122 of sub-bands within the low-band frequency range and a second group 124 of sub-bands within the high-band frequency range. have. As a non-limiting example, the low-band frequency range may be from about 0 kHz to 6.4 kHz, and the high-band frequency range may be from about 6.4 kHz to 12.8 kHz. The first group of sub-bands 124 may be provided to an analysis filter bank 202 . The analysis filter bank 202 may construct the low-band signal 212 by combining the first group 122 of sub-bands. The low-band signal 212 may be provided to the low-band coder 204 .

The low-band coder 204 may correspond to the low-band analysis module 130 of FIG. 1 . For example, the low-band coder 204 is configured to quantize the low-band signal 212 (eg, the first group of sub-bands 122 ) to generate the low-band excitation signal 144 . it might be The low-band excitation signal 144 may be provided to a nonlinear transform generator 190 .

As described with respect to FIG. 1 , the low-band excitation signal 144 is subjected to a first group of sub-bands 122 (eg, the input audio signal 102 ) using the low-band analysis module 130 . from the low-band portion of . The non-linear transform generator 190 generates a harmonically extended signal 214 (eg, a non-linear transformation) based on the low-band excitation signal 144 (eg, the first group of sub-bands 122 ). linear excitation signal). The non-linear transform generator 190 may up-sample the low-band excitation signal 144 , and process the up-sampled signal with a non-linear function to be greater than a bandwidth of the low-band excitation signal 144 . A harmonically extended signal 214 having a bandwidth may be generated. For example, in a particular embodiment, the bandwidth of the low-band excitation signal 144 may be about 0 kHz to 6.4 kHz, and the bandwidth of the harmonically extended signal 214 may be about 6.4 kHz to 16 kHz. In another particular embodiment, the bandwidth of the harmonically extended signal 214 may be higher than the bandwidth of a low-band excitation signal of the same magnitude. For example, the bandwidth of the low-band excitation signal 144 may be about 0 kHz to 6.4 kHz, and the bandwidth of the harmonically extended signal 214 may be about 6.4 kHz to 12.8 kHz. In a particular embodiment, the non-linear transform generator 190 performs an absolute-value operation or a square operation on frames (or sub-frames) of the low-band excitation signal 144 to perform a harmonically extended signal 214 . ) can also be created. The harmonically extended signal 214 may be provided to a noise combiner 206 .

The noise combiner 206 may be configured to mix the harmonically extended signal 214 with the modulated signal to generate a high-band excitation signal 216 . The modulated noise may be based on the envelope of the low-band signal 212 and white noise. The amount of modulated noise mixed with the harmonically extended signal 214 may be based on a mixing factor. The low-band coder 204 may generate information used by the noise combiner 206 to determine the mixing factor. The information includes the pitch delay in the first group of sub-bands 122 , the adaptive codebook gain associated with the first group of sub-bands 122 , the first group of sub-bands 122 and the second group of sub-bands 122 . Pitch correlations between groups 124 may also include any combination thereof, and the like. For example, if a harmonic of the low-band signal 212 corresponds to a speech signal (eg, a signal having relatively strong speech components and components such as relatively weak noise), the value of the mixing factor may increase. A smaller amount of modulated noise may be mixed with the harmonically extended signal 214 . Alternatively, if a harmonic of the low-band signal 212 corresponds to a noise-like signal (eg, a signal having relatively strong noise-like components and relatively weak speech components), the value of the mixing factor decreases and a larger amount of modulated noise may be mixed with the harmonically extended signal 214 . The high-band excitation signal 216 may be provided to a second analysis filter bank 192 .

A second filter analysis filter bank 192 converts the high-band excitation signal 216 to a third group of sub-bands 126 (eg, high-band) corresponding to the second group of sub-bands 124 . excitation signals) may be configured to filter (eg, split). Each sub-band HE1-HEN of the third group of sub-bands 126 may be provided to a corresponding parameter estimator 294a - 294c . Further, each sub-band (Hl-HN) of the second group of sub-bands 124 may be provided to a corresponding parameter estimator 294a - 294c .

Parameter estimators 294a - 294c may correspond to and operate in a substantially similar manner to parameter estimators 194 of FIG. 1 . For example, each parameter estimator 294a - 294c calculates a corresponding sub-bands in the third group of sub-bands 126 based on the metric of the corresponding sub-bands in the second group of sub-bands 124 . Adjustment parameters for sub-bands may be determined. For example, the first parameter estimator 294a calculates a second in the third group of sub-bands 126 based on a metric of the first sub-band H1 in the second group 124 of sub-bands. A first adjustment parameter (eg, an LPC adjustment parameter and/or a gain adjustment parameter) may be determined for one sub-band (HE1). For example, the first parameter estimator 294a calculates a first sub-band (HE1) in the third group of sub-bands 126 and a first sub-band in the second group of sub-bands 124 . It is also possible to determine a spectral relationship and/or an envelope relationship between (H1). To illustrate, first parameter estimator 294 performs LP analysis on the first sub-band (H1) of the second group of sub-bands 124 to obtain LPCs for the first sub-band (H1). and a residual for the first sub-band (H1). The residual for the first sub-band (H1) may be compared to the first sub-band (HE1) in the third group of sub-bands 126 , and the first parameter estimator 294 determines the number of sub-bands. determine a gain parameter such that the energy of the residual of the first sub-band H1 of the second group 124 and the energy of the first sub-band HE1 of the third group 126 of sub-bands substantially match. may be As another example, the first parameter estimator 294 performs synthesis using the first sub-band HE1 of the third group of sub-bands 126 to perform synthesis of the first group of the second group of sub-bands 124 . A synthesized version of sub-band H1 may be generated. The first parameter estimator 294 determines the gain parameter such that the energy of the first sub-band H1 of the second group of sub-bands 124 approximates the energy of the synthesized version of the first sub-band H1 . may decide In a similar manner, the second parameter estimator 294b calculates a second parameter in the third group of sub-bands 126 based on the metric of the second sub-band H2 in the second group 124 of sub-bands. A second adjustment parameter for the second sub-band (HE2) may be determined.

The adjustment parameters may be quantized by a quantizer (eg, quantizer 156 of FIG. 1 ) and transmitted as high-band side information. The third group of sub-bands 126 is also subjected to further processing (eg, gain shape adjustment processing, phase adjustment processing, etc.) by other components (not shown) of the encoder (eg, system 200 ). may be adjusted based on the adjustment parameters for .

The system 200 of FIG. 2 includes synthesized high-band signal components (eg, a third group of sub-bands 126 ) and original high-band signal components (eg, of sub-bands). The degree of correlation between the second group 124 may be improved. For example, the spectral and envelope approximation between the synthesized high-band signal components and the original high-band signal components is a sub-band by sub-band basis for the metrics of the second group of sub-bands 124 of the sub-bands. It may be performed at a “finer” level by comparing to the third group of metrics 126 . The third group of sub-bands 126 may be adjusted based on adjustment parameters resulting from the comparison, which may be transmitted to a decoder to reduce audible artifacts during high-band reconstruction of the input audio signal 102 . have.

Referring to FIG. 3 , a particular embodiment of a system 300 operable to perform high-band signal modeling is shown. The system 300 includes a first analysis filter bank 110 , an analysis filter bank 202 , a low-band coder 204 , a non-linear transform generator 190 , a second analysis filter bank 192 , N noise combiners 306a-306c, and N parameter estimators 294a-294c.

During operation of the system 300 , the harmonically extended signal 214 is provided to a second analysis filter bank 192 (as opposed to the noise combiner 206 of FIG. 2 ). The second filter analysis filter bank 192 may be configured to filter (eg, split) the harmonically extended signal 214 into a plurality of sub-bands 322 . Each sub-band of the plurality of sub-bands 322 may be provided to a corresponding noise combiner 306a - 306c . For example, a first sub-band of the plurality of sub-bands 322 may be provided to a first noise combiner 306a , wherein a second sub-band of the plurality of sub-bands 322 is a second sub-band of the plurality of sub-bands 322 . 2 may be provided to the noise combiner 306b, and so on.

Each noise combiner 306a - 306c mixes a received sub-band of the plurality of sub-bands 322 with modulated noise to mix the received sub-band of the plurality of sub-bands with a third group of sub-bands 126 (eg, a plurality of high -band excitation signals (HE1-HEN)). For example, the modulated noise may be based on the envelope of the low-band signal 212 and white noise. The amount of modulated noise mixed with each sub-band of the plurality of sub-bands 322 may be based on at least one mixing factor. In a particular embodiment, the first sub-band HE1 of the third group of sub-bands 126 is obtained by mixing the first sub-band of the plurality of sub-bands 322 based on the first mixing factor. may be generated, wherein the second sub-band HE2 of the third group of sub-bands 126 is generated by mixing the second sub-band of the plurality of sub-bands 322 based on the second mixing factor. may be created. Accordingly, multiple (eg, different) mixing factors may be used to generate the third group of sub-bands 126 .

The low-band coder 204 may generate information used by each noise combiner 306a - 306c to determine respective mixing factors. For example, the information provided to the first noise combiner 306a to determine the first mixing factor is a pitch delay, an adaptive codebook associated with a first sub-band L1 of the first group of sub-bands 122 . gain, the pitch correlation between the first sub-band (L1) of the first group of sub-bands (122) and the first sub-band (H1) of the second group of sub-bands (124), or a degree thereof It may include any combination. Similar parameters for each of the sub-bands may be used to determine mixing factors for the other noise combiners 306b, 306n. In another embodiment, each noise combiner 306a - 306n may perform mixing operations based on a common mixing factor.

As described with respect to FIG. 2 , each parameter estimator 294a - 294c calculates a third group of sub-bands 126 based on a metric of the corresponding sub-bands in the second group of sub-bands 124 . ) may determine the adjustment parameters for the corresponding sub-bands in . The adjustment parameters may be quantized by a quantizer (eg, quantizer 156 of FIG. 1 ) and transmitted as high-band side information. The third group of sub-bands 126 is also subjected to further processing (eg, gain shape adjustment processing, phase adjustment processing, etc.) by other components (not shown) of the encoder (eg, system 300 ). may be adjusted based on the adjustment parameters for .

The system 300 of FIG. 3 includes synthesized high-band signal components (eg, a third group of sub-bands 126 ) and original high-band signal components (eg, a group of sub-bands). The degree of correlation between the second group 124 may be improved. For example, the spectral and envelope approximation between the synthesized high-band signal components and the original high-band signal components is a sub-band by sub-band basis for the metrics of the second group of sub-bands 124 of the sub-bands. It may be performed at a “finer” level by comparing to the third group of metrics 126 . In addition, each sub-band (eg, high-band excitation signal) in the third group of sub-bands 126 includes a first group of sub-bands 122 and a second group of sub-bands ( 124) may be generated based on characteristics (eg, pitch values) of the corresponding sub-bands to improve signal estimation. The third group of sub-bands 126 may be adjusted based on adjustment parameters resulting from the comparison, which may be transmitted to a decoder to reduce audible artifacts during high-band reconstruction of the input audio signal 102 . have.

Referring to FIG. 4 , a particular embodiment of a system 400 operable to restore an audio signal using adjustment parameters is shown. System 400 includes a non-linear transform generator 490 , a noise combiner 406 , an analysis filter bank 492 , and N adjusters 494a - 494c . In a particular embodiment, system 400 may be incorporated into a decoding system or apparatus (eg, in a wireless telephone or codec). In other embodiments, system 400 may be integrated into a set top box, music player, video player, entertainment unit, navigation device, communication device, PDA, fixed location data unit, or computer.

The non-linear transform generator 490 is configured to generate a harmonically extended signal 414 (e.g., based on the low-band excitation signal 144 received as part of the low-band bit stream 142 in the bit stream 199) , a non-linear excitation signal). The harmonically extended signal 414 may correspond to the reconstructed version 214 of the harmonically extended signal of FIGS. For example, the non-linear transform generator 490 may operate in a substantially similar manner to the non-linear transform generator 190 of FIGS. In an exemplary embodiment, the harmonically extended signal 414 may be provided to the noise combiner 406 in a manner similar to that described with respect to FIG. 2 . In another particular embodiment, the harmonically extended signal 414 may be provided to the analysis filter bank 492 in a manner similar to that described with respect to FIG. 3 .

A noise combiner 406 is configured to receive the low-band bit stream 142 and produce a mixing factor, as described for noise combiner 206 of FIG. 2 or noise combiners 306a - 306c of FIG. 3 . may be Alternatively, the noise combiner 406 may receive the high-band side information 172 that includes the mixing factor generated at the encoder (systems 100 - 300 of FIGS. 1-3 ). In the exemplary embodiment, the noise combiner 406 mixes the transformed low-band excitation signal 414 with modulated noise to generate the high-band excitation signal 416 (eg, the high-band excitation signal 416 of FIG. 2 ) based on the mixing factor. - a reconstructed version of the band excitation signal 216). For example, the noise combiner 406 may operate in a substantially similar manner to the noise combiner 206 of FIG. 2 . In an exemplary embodiment, the high-band excitation signal 416 may be provided to an analysis filter bank 492 .

In an exemplary embodiment, the analysis filter bank 492 converts the high-band excitation signal 416 to a group of high-band excitation sub-bands 426 (eg, the second of the sub-bands of FIGS. 1-3 ). The reconstructed version of the second group of 3 groups 126) may be configured to filter (eg, split). For example, analysis filter bank 492 may operate in a substantially similar manner as second analysis filter bank 192 as described with respect to FIG. 2 . The group 426 of high-band excitation sub-bands may be provided to a corresponding adjuster 494a - 494c .

In another embodiment, the analysis filter bank 492 converts the harmonically extended signal 414 into a plurality of sub-bands (not shown) in a manner similar to the second analysis filter bank 192 as described with respect to FIG. 3 . It may be configured to filter by . In this embodiment, multiple noise combiners (not shown) combine each sub-band of the plurality of sub-bands with the modulated noise (based on mixing factors transmitted as high-band side information). The group 426 of the high-band excitation sub-bands may be generated in a similar manner to the noise combiners 394a - 394c of 3 . Each sub-band of the groups of high-band excitation sub-bands 426 may be provided to a corresponding adjuster 494a - 494c .

Each adjuster 494a - 494c may receive a corresponding adjustment parameter generated by parameter estimators 194 of FIG. 1 , such as high-band side information 172 . Each coordinator 494a - 494c may also receive a corresponding sub-band of group 426 of high-band excitation sub-bands. The adjusters 494a - 494c may be configured to generate an adjusted group 424 of high-band excitation sub-bands based on the adjustment parameters. The adjusted group 424 of high-band excitation sub-bands is transferred to other components (not shown) of the system 400 for further processing (eg, LP synthesis, gain shape adjustment processing, phase adjustment processing, etc.) may be provided to reconstruct the second group 124 of the sub-bands of FIGS. 1-3 .

The system 400 of FIG. 4 uses the low-band bit stream 142 of FIG. 1 and the adjustment parameters (eg, the high-band side information 172 of FIG. 1 ) to a second group of sub-bands. (124) may be restored. Using the adjustment parameters may improve the accuracy of the reconstruction (eg, produce a finely tuned reconstruction) by performing adjustment of the high-band excitation signal 416 on a sub-band basis.

Referring to FIG. 5 , a flowchart of a particular embodiment of a method 500 for performing high-band signal modeling is shown. As an illustrative example, the method 500 may be performed by one or more of the systems 100 - 300 of FIGS. 1-3 .

The method 500 may include, at a speech encoder, filtering the audio signal into a first group of sub-bands within a first frequency range and a second group of sub-bands within a second frequency range, at 502 . . For example, referring to FIG. 1 , a first analysis filter bank 110 may convert an input audio signal 102 into a first group of sub-bands 122 within a first frequency range and a sub-band within a second frequency range. may filter to a second group 124 of The first frequency range may be lower than the second frequency range.

A harmonically extended signal may be generated based on the first group of sub-bands, at 504 . For example, referring to FIGS. 2-3 , the analysis filter bank 202 may generate the low-band signal 212 by combining the first group 122 of sub-bands, wherein the low-band coder 204 may encode the low-band signal 212 to generate a low-band excitation signal 144 . The low-band excitation signal 144 may be provided to a non-linear transform generator 407 . The non-linear transform generator 190 generates a harmonically extended signal 214 (eg, a non-linear transformation) based on the low-band excitation signal 144 (eg, the first group of sub-bands 122 ). The low-band excitation signal 144 may be up-sampled to generate a linear excitation signal).

A third group of sub-bands may be generated based, at least in part, on the harmonically extended signal, at 506 . For example, referring to FIG. 2 , the harmonically extended signal 214 may be mixed with modulated noise to generate a high-band excitation signal 216 . A second filter analysis filter bank 192 converts the high-band excitation signal 216 to a third group of sub-bands 126 (eg, high-band) corresponding to the second group of sub-bands 124 . excitation signals) may be filtered (eg, split). Alternatively, with reference to FIG. 3 , the harmonically extended signal 214 is provided to a second analysis filter bank 192 . The second filter analysis filter bank 192 may filter (eg, split) the harmonically extended signal 214 into a plurality of sub-bands 322 . Each sub-band of the plurality of sub-bands 322 may be provided to a corresponding noise combiner 306a - 306c . For example, a first sub-band of the plurality of sub-bands 322 may be provided to a first noise combiner 306a , wherein a second sub-band of the plurality of sub-bands 322 is a second sub-band of the plurality of sub-bands 322 . 2 may be provided to the noise combiner 306b, and so on. Each noise combiner 306a - 306c may mix the received sub-band of the plurality of sub-bands 322 with modulated noise to produce a third group of sub-bands 126 .

At 508 , a first adjustment parameter for a first sub-band in the third group of sub-bands may be determined, or a second adjustment parameter for a second sub-band in the third group of sub-bands may be determined. may be For example, referring to FIGS. 2-3 , the first parameter estimator 294a calculates a metric (eg, signal energy) of the corresponding sub-band H1 in the second group of sub-bands 124 . , residual energy, LP coefficients, etc.) a first adjustment parameter (eg, LPC adjustment parameter and/or gain for the first sub-band HE1 in the third group of sub-bands 126 ) adjustment parameters). The first parameter estimator 294a may calculate a first gain factor (eg, a first adjustment parameter) according to a relationship between the first sub-band HE1 and the first sub-band H1 . The gain factor may correspond to the difference (or ratio) between the energies of the sub-bands H1 , HE1 over a frame or some portion of the frame. In a similar manner, other parameter estimators 294b - 294c calculate the metric (eg, signal energy, residual energy, LP coefficients) of the second sub-band H2 in the second group of sub-bands 124 . etc.), determine a second adjustment parameter for the second sub-band HE2 in the third group of sub-bands 126 .

The method 500 of FIG. 5 includes synthesized high-band signal components (eg, a third group of sub-bands 126 ) and original high-band signal components (eg, of sub-bands). The degree of correlation between the second group 124 may be improved. For example, the spectral and envelope approximation between the synthesized high-band signal components and the original high-band signal components is a sub-band by sub-band basis for the metrics of the second group of sub-bands 124 of the sub-bands. It may be performed at a “finer” level by comparing to the third group of metrics 126 . The third group of sub-bands 126 may be adjusted based on adjustment parameters resulting from the comparison, which may be transmitted to a decoder to reduce audible artifacts during high-band reconstruction of the input audio signal 102 . have.

Referring to FIG. 6 , a flowchart of a particular embodiment of a method 600 of reconstructing an audio signal using adjustment parameters is shown. As an illustrative example, the method 600 may be performed by the system 400 of FIG. 4 .

Method 600 includes generating a harmonically extended signal based on a low-band excitation signal received from a speech encoder, at 602 . For example, referring to FIG. 4 , a low-band excitation signal 444 is provided to a non-linear transform generator 490 based on the low-band excitation signal 444 to a harmonically extended signal 414 (eg, , a non-linear excitation signal).

The group of high-band excitation sub-bands may be generated based, at least in part, on the harmonically extended signal, at 606 . For example, referring to FIG. 4 , the noise combiner 406 may determine a mixing factor based on a pitch delay between bands, an adaptive codebook gain, and/or a pitch correlation between bands as described with respect to FIG. 4 , or , may receive high-band side information 172 including a mixing factor generated at an encoder (eg, systems 100 - 300 of FIGS. 1-3 ). A noise combiner 406 mixes the transformed low-band excitation signal 414 with modulated noise and based on the mixing factor a high-band excitation signal 416 (eg, the high-band excitation signal 216 of FIG. 2 ) ) ) can also be created. Analysis filter bank 492 converts high-band excitation signal 416 into group of high-band excitation sub-bands 426 (eg, third group of sub-bands 126 of FIGS. 1-3 ). The second group of reconstructed versions) may be filtered (eg, split).

The group of high-band excitation sub-bands may be adjusted based on adjustment parameters received from the speech encoder, at 608 . For example, referring to FIG. 4 , each adjuster 494a - 494c may receive a corresponding adjustment parameter generated by parameter estimators 194 of FIG. 1 , such as high-band side information 172 . have. Each coordinator 494a - 494c may also receive a corresponding sub-band of group 426 of high-band excitation sub-bands. The adjusters 494a - 494c may generate an adjusted group 424 of high-band excitation sub-bands based on the adjustment parameters. The adjusted group 424 of high-band excitation sub-bands is provided to another component (not shown) of the system 400 for further processing (eg, gain shape adjustment processing, phase adjustment processing, etc.) in FIG. 1 . The second group 124 of the sub-bands of FIG. 3 may be reconstructed.

The method 600 of FIG. 6 uses the low-band bit stream 142 of FIG. 1 and the adjustment parameters (eg, the high-band side information 172 of FIG. 1 ) to a second group of sub-bands. (124) may be restored. Using the adjustment parameters may improve the accuracy of the reconstruction (eg, produce a finely tuned reconstruction) by performing adjustment of the high-band excitation signal 416 on a sub-band basis.

In certain embodiments, the methods 500 , 600 of FIGS. 5 and 6 may be applied to hardware (eg, FPGA device, ASIC) of a processing unit such as a central processing unit (CPU), DSP, or controller. etc.), via a firmware device, or any combination thereof. As an example, the methods 500 , 600 of FIGS. 5 and 6 may be performed by a processor executing the instructions, as described with respect to FIG. 7 .

Referring to FIG. 7 , a block diagram of a particular illustrative embodiment of a wireless communication device is shown and generally designated 700 . Device 700 includes a processor 710 (eg, CPU) coupled to memory 732 . Memory 732 may be used by processor 710 and/or codec 734 to perform the methods and processes disclosed herein, such as one or both of methods 500 , 600 of FIGS. 5 and 6 . executable instructions 760 .

In a particular embodiment, the codec 734 may include an encoding system 782 and a decoding system 784 . In a particular embodiment, encoding system 782 includes one or more components of systems 100 - 300 of FIGS. 1-3 . For example, encoding system 782 may perform encoding operations associated with systems 100 - 300 of FIGS. 1-3 and method 500 of FIG. 5 . In a particular embodiment, the decoding system 784 includes one or more components of the system 400 of FIG. 4 . For example, the decoding system 784 may perform decoding operations associated with the system 400 of FIG. 4 and the method 600 of FIG. 6 .

The encoding system 782 and/or the decoding system 784 may be implemented via dedicated hardware (eg, circuitry), by a processor executing instructions to perform one or more tasks, or a combination thereof. For example, memory 732 or memory 790 in codec 734 may include random access memory (RAM), magnetoresistive random access memory (MRAM), spin-torque transfer MRAM (STT-MRAM), flash memory, read - Dedicated Memory (ROM), Programmable Read-Only Memory (PROM), Erasable Programmable Read-Only Memory (EPROM), Electrically Erasable Programmable Read-Only Memory (EEPROM), Registers, Hard Disk, Removal possible disk, or a memory device such as a compact disk read-only memory (CD-ROM). The memory device, when executed by a computer (eg, processor in codec 734 and/or processor 710 ), causes the computer to: one of methods 500 , 600 of FIGS. 5 and 6 . instructions (eg, instructions 760 or instructions 785 ) to perform at least a portion of By way of example, memory 732 or memory 790 in codec 734 can cause a computer to, when executed by a computer (eg, processor and/or processor 710 in codec 734 ). , a non-transitory computer comprising instructions (eg, instructions 760 or instructions 795, respectively) to perform at least a portion of one of the methods 500 and 600 of FIGS. 5 and 6 , respectively. - It may be a readable medium.

The device 700 may also include a codec 734 and a DSP 796 coupled to the processor 710 . In a particular embodiment, the DSP 796 may include an encoding system 797 and a decoding system 798 . In a particular embodiment, encoding system 797 includes one or more components of systems 100 - 300 of FIGS. 1-3 . For example, encoding system 797 may perform encoding operations associated with systems 100 - 300 of FIGS. 1-3 and method 500 of FIG. 5 . In a particular embodiment, the decoding system 798 may include one or more components of the system 400 of FIG. 4 . For example, the decoding system 798 may perform decoding operations associated with the system 400 of FIG. 4 and the method 600 of FIG. 6 .

7 also shows a display controller 726 coupled to a processor 710 and a display 728 . The codec 734 may be coupled to the processor 710 , as shown. A speaker 736 and a microphone 738 can be coupled to the codec 734 . For example, the microphone 734 may generate the input audio signal 102 of FIG. 1 , and the codec 734 may generate the output bit stream 199 for transmission to a receiver based on the input audio signal 102 . can also create For example, the output bit stream 199 may be transmitted to a receiver via a processor 710 , a wireless controller 740 , and an antenna 742 . As another example, the speaker 736 may be used to output a signal reconstructed by the codec 734 from the output bit stream 199 of FIG. 1 , the output bit stream 199 from a transmitter (eg, via a radio controller 740 and an antenna 742).

In a particular embodiment, processor 710 , display controller 726 , memory 732 , codec 734 , and wireless controller 740 are system-in-package or system-on-chip devices (eg, , mobile station modem (MSM)) 722 . In a particular embodiment, an input device 730 , such as a touchscreen and/or keypad, and a power supply 744 are coupled to the system-on-chip device 722 . Further, in a particular embodiment, as shown in FIG. 7 , a display 728 , an input device 730 , a speaker 736 , a microphone 738 , a wireless antenna 742 , and a power supply 744 are provided in the system - is external to the on-chip device 722 . However, each of the display 728 , the input device 730 , the speaker 736 , the microphone 738 , the antenna 742 , and the power supply 744 is an interface or controller of the system-on-a-chip device 722 . It can be connected to a component.

In conjunction with the described embodiments, a first apparatus is disclosed comprising means for filtering an audio signal into a first group of sub-bands within a first frequency range and a second group of sub-bands within a second frequency range. do. For example, the means for filtering the audio signal may include the first analysis filter bank 110 of FIGS. 1-3 , the encoding system 782 of FIG. 7 , the encoding system 797 of FIG. 7 , configured to filter the audio signal. may include one or more devices (eg, a processor executing instructions in a non-transitory computer-readable medium), or any combination thereof.

The first apparatus may also include means for generating a harmonically extended signal based on the first group of sub-bands. For example, the means for generating the harmonically extended signal includes the low-band analysis module 130 of FIG. 1 and its components, the non-linear transformation generator 190 of FIGS. 1-3 , FIGS. 2 and 3 . synthesis filter bank 202 of FIG. 2 and 3 , low-band coder 204 of FIG. 2 and 3 , encoding system 782 of FIG. 7 , encoding system 797 of FIG. devices (eg, a processor executing instructions in a non-transitory computer-readable storage medium), or any combination thereof.

The first apparatus may also include means for generating a third group of sub-bands based at least in part on the harmonically extended signal. For example, the means for generating the third group of sub-bands includes the high-band analysis module 150 of FIG. 1 and its components, the second analysis filter bank 192 of FIGS. 1-3 , FIG. 2 . noise combiner 206 of FIG. 3 , noise combiners 306a - 306c of FIG. 3 , encoding system 782 of FIG. 7 , one or more devices configured to generate the third group of sub-bands (eg, non-transient processor for executing instructions in a computer-readable storage medium), or any combination thereof.

The first apparatus also includes means for determining a first adjustment parameter for a first sub-band in the third group of sub-bands or a second adjustment parameter for a second sub-band in the third group of sub-bands may include For example, the means for determining the first and second adjustment parameters may include the parameter estimators 194 of FIG. 1 , the parameter estimators 294a-294c of FIG. 2 , the encoding system 782 of FIG. encoding system 797 , one or more devices configured to determine the first and second adjustment parameters (eg, a processor executing instructions in a non-transitory computer-readable storage medium), or any combination thereof may be

In conjunction with the described embodiments, a second apparatus is disclosed comprising means for generating a harmonically extended signal based on a low-band excitation signal received from a speech encoder. For example, the means for generating the harmonically extended signal includes the non-linear transform generator 490 of FIG. 4 , the decoding system 784 of FIG. 7 , the decoding system 798 of FIG. 7 , to generate the harmonically extended signal. may include one or more devices configured (eg, a processor executing instructions in a non-transitory computer-readable storage medium), or any combination thereof.

The second apparatus may also include means for generating the group of high-band excitation sub-bands based at least in part on the harmonically extended signal. For example, the means for generating the group of high-band excitation sub-bands includes the noise combiner 406 of FIG. 4 , the analysis filter bank 492 of FIG. 4 , the decoding system 784 of FIG. 7 , the decoding of FIG. may include system 798 , one or more devices configured to generate the group of high-band excitation signals (eg, a processor executing instructions in a non-transitory computer-readable storage medium), or any combination thereof. have.

The second apparatus may also include means for adjusting the group of high-band excitation sub-bands based on the adjustment parameters received from the speech encoder. For example, the means for adjusting the group of high-band excitation sub-bands may include the adjusters 494a-494c of FIG. 4 , the decoding system 784 of FIG. 7 , the decoding system 798 of FIG. 7 , the high-band It may include one or more devices configured to coordinate the group of sub-bands (eg, a processor executing instructions in a non-transitory computer-readable storage medium), or any combination thereof.

Those of ordinary skill in the art will appreciate that the various illustrative logical blocks, configurations, modules, circuits, and algorithm steps described in connection with the embodiments disclosed herein are executed by a processing device, such as electronic hardware, a hardware processor, computer software; or as combinations of both. Various illustrative components, blocks, configurations, modules, circuits, and steps have been described above generally in terms of their functionality. Whether such functionality is implemented as hardware or software depends upon the particular application and design constraints imposed on the overall system. Skilled artisans may implement the described functions in varying ways for each particular applications, but such implementation decisions should not be interpreted as causing a departure from the scope of the present disclosure.

The steps of a method or algorithm described in connection with the embodiments disclosed herein may be implemented directly in hardware, in a software module executed by a processor, or in a combination of both. Software modules include random access memory (RAM), magnetoresistive random access memory (MRAM), spin-torque transfer MRAM (STT-MRAM), flash memory, read-only memory (ROM), programmable read-only memory (PROM). ), erasable programmable read-only memory (EPROM), electrically erasable programmable read-only memory (EEPROM), registers, hard disk, removable disk, or compact disk read-only memory (CD-ROM) It may be in a memory device such as An exemplary memory device is coupled to the processor such that the processor can read information from, and write information to, the memory device. Alternatively, the memory device may be integrated into the processor. The processor and storage medium may be within the ASIC. The ASIC may be within a computing device or user terminal. In the alternative, the processor and storage medium may be as separate components in the computing device or user terminal.

The previous description of the disclosed embodiments is provided to enable any person skilled in the art to make or use the disclosed embodiments. Various modifications to these embodiments will be readily apparent to those skilled in the art, and the generic principles defined herein may be applied to other embodiments without departing from the scope of the disclosure. Accordingly, the present disclosure is not intended to be limited to the exemplary embodiments shown herein, but is to be accorded the widest possible scope consistent with the principles and novel features defined by the following claims.

Claims (10)

In the speech encoder, the audio signal is divided into a first group of sub-band signals (L1, L2, ..., LM) in a first frequency range and a second group of sub-band signals (H1, H2, ..., HN) filtering;
performing a linear prediction analysis to generate a first residual signal of a first sub-band (H1) in the second group of sub-bands;
performing a linear prediction analysis to generate a second residual signal of a second sub-band (H2) in the second group of sub-bands;
combining the first group of sub-band signals to produce a low-band signal and quantizing the low-band signal to produce a low-band excitation signal;
generating a harmonically extended signal based on the low-band excitation signal and a non-linear processing function;
generating a third group of sub-band signals (HE1, HE2, ..., HEN) based at least in part on the harmonically extended signal, wherein the third group of sub-bands are sub-bands of the second group - generating the third group of sub-band signals, corresponding to the bands; and
a first adjustment parameter for a first sub-band signal HE1 in the third group of sub-band signals and a second sub-band signal HE2 in the third group of sub-band signals determining a second adjustment parameter for , the first adjustment parameter such that the energy of the first sub-band signal (HE1) of the third group of sub-band signals is equal to the energy of the first residual signal adjust a gain, and the second adjustment parameter adjusts the gain so that the energy of the second sub-band signal (HE2) of the third group of sub-band signals is equal to the energy of the second residual signal A method for modeling a high-band signal comprising determining. The method of claim 1,
and the first adjustment parameter and the second adjustment parameter correspond to linear prediction coefficient adjustment parameters. The method of claim 1,
and inserting the first adjustment parameter and the second adjustment parameter into the encoded version of the audio signal to enable adjustment during reconstruction of the audio signal from the encoded version of the audio signal. method for modeling The method of claim 1,
The generating of the third group of sub-band signals comprises:
mixing the harmonically extended signal with modulated noise to produce a high-band excitation signal, wherein the modulated noise and the harmonically extended signal are based on a mixing factor; and
and filtering the high-band excitation signal into the third group of sub-band signals. 5. The method of claim 4,
The mixing factor may be a pitch delay, an adaptive codebook gain associated with the first group of sub-band signals, or a pitch correlation between the first group of sub-band signals and the second group of sub-band signals. A method for modeling a high-band signal, which is determined based on at least one of: The method of claim 1,
The generating of the third group of sub-band signals comprises:
filtering the harmonically extended signal into a plurality of sub-band signals; and
mixing each of the sub-band signals of the plurality of sub-band signals with modulated noise to generate a plurality of high-band excitation signals, the plurality of high-band excitation signals comprising the third group of sub-bands; and mixing each of the sub-band signals corresponding to signals. 7. The method of claim 6,
the modulated noise and a first sub-band signal of the plurality of sub-band signals are mixed based on a first mixing factor, and a second sub-band signal of the modulated noise and the plurality of sub-band signals is A method for modeling a high-band signal, which is mixed based on a second mixing factor. The audio signal is divided into a first group of sub-band signals (L1, L2,..., LM) in a first frequency range and a second group of sub-band signals (H1, H2,..., HN) in a second frequency range. means to filter by;
means for performing a linear prediction analysis to generate a first residual signal of a first sub-band (H1) in the second group of sub-bands;
means for performing a linear prediction analysis to generate a second residual signal of a second sub-band (H2) in the second group of sub-bands;
means for combining the first group of sub-band signals to produce a low-band signal and quantizing the low-band signal to produce a low-band excitation signal;
means for generating a harmonically extended signal based on the low-band excitation signal and a non-linear processing function;
means for generating a third group of sub-band signals (HE1, HE2, ..., HEN) based at least in part on the harmonically extended signal, wherein the third group of sub-bands are sub-bands of the second group - means for generating the third group of sub-band signals corresponding to bands; and
a first adjustment parameter for a first sub-band signal in the third group of sub-band signals and a second adjustment parameter for a second sub-band signal in the third group of sub-band signals means for determining, wherein the first adjustment parameter adjusts a gain such that an energy of the first sub-band signal of the third group of sub-band signals is equal to an energy of the first residual signal, and the second adjustment The parameter comprises means for determining, adjusting a gain such that the energy of the second sub-band signal and the energy of the second residual signal of the sub-band signals of the third group are equal. A device for modeling. 9. The method of claim 8,
and the first adjustment parameter and the second adjustment parameter correspond to linear prediction coefficient adjustment parameters. A non-transitory computer readable medium comprising instructions that, when executed by a processor in a speech encoder, cause the processor to execute the method of any one of claims 1-7.

Images